Scalable Hybrid Large-Scale dc-ac Grid Analysis Methods (Phase I)

Suman Debnath, Marcelo Elizondo, Jiazi Zhang, J. S. Siva Prasad, Quan Nguyen, Jinho Kim, Michael Abdelmalak, Sameer Nekkalapu, Bharat Vyakaranam, Ahmad Tbaileh, Akrivi Liakopoulou, Shuan Dong, Ningchao Gao, P. Brown, Randika Bandara Wijekoon, Leonardo Rese, Hongfei Sun, Xiaofei Wang, Mingjian Tuo, Joshua Novacheck, Harry M. Hughes, M.f. Martinez-Montejano, Misael Martinez Montejano2025 · DOI

Abstract: The goals of the project included the development of characterization methods and tools to evaluate reliability, transient stability, and economics of large-scale dc architectures in ac grids.

Introduction

As the North American power grid undergoes a fundamental transformation toward a cleaner, more resilient future, the integration of High-Voltage Direct Current (HVDC) and Multi-Terminal DC (MTdc) systems has become a critical priority. However, the complexity of these systems presents a significant challenge for grid planners. Traditional analysis tools, designed for a grid dominated by massive rotating machines, are increasingly inadequate for capturing the fast, non-linear dynamics of modern power electronics.

This report, a collaborative effort between Oak Ridge National Laboratory (ORNL), Pacific Northwest National Laboratory (PNNL), and the National Renewable Energy Laboratory (NREL), details the first phase of a project aimed at bridging the gap between high-level grid planning and the complex physics of power electronics. The work introduces a multi-scale modeling framework that spans from long-term economic planning to microsecond-level transient simulations.

The Simulation Challenge: The "Goldilocks" Problem

The power industry currently faces a significant hurdle in simulation fidelity. Grid planners typically rely on two types of models, neither of which is perfectly suited for large-scale hybrid AC-DC systems:

  1. Transient Stability (TS) Models: These models are computationally efficient and can simulate continental-scale grids with over 100,000 buses. However, they are too simplified to capture the fast, non-linear switching physics and high-frequency phenomena of modern DC converters.
  2. Electromagnetic Transient (EMT) Models: These models capture every detail of power electronic switching but are so computationally intensive that they are generally limited to small, isolated portions of the grid.

As the grid becomes increasingly dominated by power electronics, understanding phenomena like transient overvoltages (TOV) and fast frequency swings—which TS models often miss—is vital. This project addresses the lack of integrated tools capable of evaluating the reliability, stability, and economics of massive hybrid AC-DC architectures simultaneously.

Methodology: A Multi-Scale Modeling Framework

The researchers developed a "vertical" integration strategy involving four distinct modeling tiers to ensure that high-level planning is grounded in physical reality:

  • Capacity Expansion (ReEDS): Used to identify optimal locations for building DC lines over multi-decade horizons.
  • Production Cost (PLEXOS): Employed to simulate hourly grid operations and evaluate the economic impact of MTdc integration.
  • Transient Stability (PSS/E): Used for modeling large-scale electromechanical dynamics across the entire interconnection.
  • Electromagnetic Transients (PSCAD): Utilized for high-fidelity modeling of converters, breakers, and local controls.

The core innovation of this project is the EMT-TS Co-Simulation setup. Using a "buffer zone" methodology, the team identified sensitive AC buses surrounding DC terminals through reactive power injection studies. These critical buffer zones are simulated in high-fidelity EMT, while the remainder of the 110,000-bus North American grid is simulated in the faster TS domain. This approach enables "physics-aware" planning at a continental scale.

Key Insights and Grid Dynamics

The study revealed several critical insights into the behavior of hybrid grids:

  • "Invisible" Faults: Standard TS simulations were found to fail in capturing converter tripping caused by transient overvoltages. These events are effectively invisible to traditional tools but could trigger cascading failures in a real-world hybrid grid. The co-simulation framework brings these risks to light.
  • MTdc as a Grid Catalyst: Beyond simple power transmission, MTdc systems can act as a "firewall" between interconnections. The study demonstrates that DC converters decouple AC and DC power, allowing the DC system to maintain stability even when the AC side experiences severe faults.
  • Black Start Acceleration: Voltage Source Converter (VSC) based HVDC systems can serve as black start units. Because they can independently create their own voltage and frequency reference, they can provide the "cranking power" needed to restart traditional thermal power plants significantly faster than conventional diesel generators.

Key Results

Phase I of the project yielded several landmark results:

  • 10-Terminal MTdc Design: The team successfully designed and simulated a radial 10-terminal MTdc system spanning the United States. This design utilized a mixed monopole and bipole architecture to balance cost and power requirements, with stations ranging from 1 GW to 6 GW.
  • Blackout Recovery: In a reconstruction of the 2011 Southwest Blackout, the team demonstrated that VSC-assisted black start could restore 10.74% more load (approximately 2,002 MWh) compared to traditional restoration methods.
  • Computational Feasibility: The co-simulation of the 110,000+ bus model achieved an average CPU time of roughly 56 minutes per second of simulation time. While computationally intensive, this represents a significant step toward making continental-scale EMT analysis a practical reality for grid planners.

Limitations and Future Work

While Phase I established a robust framework, the authors identified several areas for future research. As the number of Inverter-Based Resources (IBRs) like solar and wind grows, future work must address the complex interactions between HVDC stations and smaller inverters across various timescales. Additionally, there is a need for more sophisticated fault detection and coordination algorithms across symmetric and asymmetric DC systems. Finally, as the grid scales, further research into parallel computing and model reduction will be necessary to overcome remaining computational bottlenecks.

Summary updated

Abstract

The goals of the project included the development of characterization methods and tools to evaluate reliability, transient stability, and economics of large-scale dc architectures in ac grids. | OSTI.GOV
BibTeX
@misc{debnath2025scalable,
  author = {Suman Debnath and Marcelo Elizondo and Jiazi Zhang and J. S. Siva Prasad and Quan Nguyen and Jinho Kim and Michael Abdelmalak and Sameer Nekkalapu and Bharat Vyakaranam and Ahmad Tbaileh and Akrivi Liakopoulou and Shuan Dong and Ningchao Gao and P. Brown and Randika Bandara Wijekoon and Leonardo Rese and Hongfei Sun and Xiaofei Wang and Mingjian Tuo and Joshua Novacheck and Harry M. Hughes and M.F. Martinez-Montejano and Misael Martinez Montejano},
  title = {{Scalable Hybrid Large-Scale dc-ac Grid Analysis Methods (Phase I)}},
  year = {2025},
  doi = {10.2172/3001758},
}

Loading related content...

Figure